introvoyz041/chemicals2

该数据集是一个化学分子结构数据集，包含4,558,282个训练样本，总大小约2.32 GB。数据特征包括样本标识（sample）、国际化合物标识（InChI）、自引用嵌入字符串（SELFIES）、原始简化分子线性输入规范（SMILES_ORIGINAL）以及数据来源分割信息（source_split）。这些特征用于表示化学分子的不同编码格式，适用于化学信息学、药物发现和机器学习任务。

This dataset is a chemical molecular structure dataset containing 4,558,282 training samples with a total size of approximately 2.32 GB. Its data features include sample identifier (sample), International Chemical Identifier (InChI), Self-Referencing Embedded Strings (SELFIES), original Simplified Molecular Input Line Entry System (SMILES_ORIGINAL), and data source split information (source_split). These features represent different encoding formats of chemical molecules, and are applicable to chemoinformatics, drug discovery and machine learning tasks.